Hepatitis E virus seroprevalence among the general population in a livestock-dense area in the Netherlands: a cross-sectional population-based serological survey

Hepatitis E virus (HEV) is a major cause of enterically transmitted hepatitis, especially in developing countries. However, recent studies show that autochthonous HEV infections are an increasing public health concern in industrialised countries [1–12].

The most common clinical presentation of HEV infection is acute hepatitis. Usually this is a self-limiting disease, with jaundice as the most common symptom [4, 13–15]. In patients with pre-existing liver disease and immunosuppressed patients, the clinical course of disease is often more severe and in developing countries excess mortality is seen in pregnant women [8, 10, 14–16]. However, most HEV infections in industrialised countries are either unrecognised or asymptomatic [1, 14, 16, 17]. This is supported by recent serological studies in several European countries, showing that the anti-HEV IgG seroprevalence in healthy blood donors varied between 7 and 52%, depending on both geographical area and the antibody assay used [17–22].

HEV is a non-enveloped single-stranded RNA virus belonging to the Hepeviridae family [16, 17, 23]. Four major genotypes, within species Orthohepevirus A, can infect humans [10, 15, 23, 24]. The genotypes 1 and 2 are restricted to human beings, and the genotypes 3 and 4 can infect both human beings and mammals. HEV-1 and HEV-2 are endemic in developing countries, leading to sporadic cases as well as large outbreaks. HEV-3 and HEV-4 are responsible for an increasing number of autochthonous hepatitis E infections worldwide [4, 10, 16, 17, 25]. The more recently discovered HEV genotypes 5 and 6 are so far only detected in animals, while genotype 7 is also described in humans [10, 26–29].

In contrast to HEV-1 and HEV-2, for which the faecal-oral transmission route via contaminated water has been confirmed, the transmission routes of HEV-3 and HEV-4 are largely unclear and the exact source of infection remains unknown for the majority of patients [4, 17, 27, 28]. The increasing trend in the number of autochthonous cases in Western European countries stresses the importance to get insight in the route of infection, enabling the implementation of control measures against human HEV infections [27, 28, 30].

It is hypothesized that HEV-3 and HEV-4 have a zoonotic origin, supported by the high prevalence of HEV-3 and HEV-4 among domestic pigs and wild boars [23, 26, 31]. In the Netherlands, the prevalence rate of HEV-3 in the domestic pig population is estimated at 55%, and domestic pigs may therefore be an important reservoir for human HEV infections [28, 32]. Furthermore, the frequent detection of HEV in pork products and the high similarity between porcine and human HEV sequences suggest zoonotic transmission of HEV-3 and HEV-4 worldwide [33–36]. Zoonotic food-borne transmission has been proven by the identification of identical nucleotide sequences in autochthonous HEV patients and in leftover portions of consumed contaminated food [37–40]. But zoonotic HEV transmission might also occur through exposure to contaminated environments. Infected pigs, which are generally asymptomatic, can excrete large amounts of virus via their faeces in the environment. This may lead to human infections [26–28, 41]. The high anti-HEV serum antibody rates in humans with occupational contact with pigs, like farmers and veterinarians, provide indirect evidence for this route of transmission [42–47]. However, environmental exposure to HEV might also pose a risk for of neighbouring residents of livestock farming, as was the case during the major Dutch Q-fever outbreak in 2007–2010 [48].

In the Netherlands there is an ongoing debate on the environmental health risks as a result of (intensive) livestock farming in areas highly populated with both livestock and people. At a surface of approximately 34.000 km² almost 17 million people are living together with 75 million chickens, 7 million pigs, 4 million cattle and 1.5 million goats and sheep. Within a large study on the health of people living in the vicinity of livestock farming, we performed a serological survey in order to assess the HEV seroprevalence among non-farming adults from the general population in a livestock-dense area in the south of the Netherlands. Furthermore, we investigated whether living close to pig farms is a risk factor for HEV seropositivity, focusing on the role of environmental transmission of HEV from pig farms to residents living in the vicinity of these farms.

In 2014 the Livestock Farming and Neighbouring Residents’ Health study (Dutch acronym: VGO) started, aiming to assess the relationship between livestock farming and a number of health outcomes in people living in the eastern part of the province of Noord-Brabant and the northern part of the province of Limburg. This part of the Netherlands is a relatively highly populated rural area with a high density of livestock farms. Within the VGO study a cross-sectional health study was performed, investigating three categories of health effects: 1) respiratory health effects, 2) livestock-associated infections, and 3) carriage of resistant microorganisms. Furthermore, air sampling was conducted in order to assess to what extent residents living in the vicinity of livestock farms were exposed to emissions of these farms, including endotoxin amount, PM10 and DNA of Staphylococcus aureus and E. coli. Participants of the cross-sectional health study provided blood samples in order to assess (previous) exposure to several livestock-farming associated pathogens, including HEV.

Study population

The study population was selected in a two-step procedure. First, a questionnaire survey was conducted among 14,163 unrelated adults from the general population (aged 18–70 years) living in the study area. Participants were recruited via their general practitioner (GP). In the Netherlands each individual is registered with just one general practice, mostly close to residential address. Pre-defined inclusion criteria have been described previously by Borlée et al. [49]. Second, questionnaire participants who gave consent to be contacted for further studies, and who were not working or living on a farm, were eligible for the serological survey (n = 8,714) [50]. Based on their home addresses, twelve temporary study centres were established. Figure 1 shows the location of the temporary study centres, and pig density in the Netherlands in 2014. All participants living within a distance of approximately 10 km of a temporary study centre (n = 7,180) were invited to the nearest centre for a medical examination, including blood sampling for serological analyses against livestock-associated pathogens. Furthermore, they were asked to fill in a questionnaire comprising items on symptoms and diseases, home characteristics, smoking habits, education, profession, leisure activities, dietary habits (including the consumption of pork meat), and animal contact, including exposure to animal farm environment during childhood. Data were collected between 10 March 2014 and 27 February 2015. Patients’ privacy was ensured as described earlier [49–51]. In short, medical information and address records were kept separated at all times by using a Trusted Third Party (Stichting Informatie Voorziening Zorg, Houten). The VGO study protocol was approved by the Medical Ethical Committee of the University Medical Centre Utrecht. All participants signed informed consent.

Positive and borderline positive anti-HEV IgG ratio was found for respectively 666 (27.5%) and 29 (1.2%) of the 2,422 participants. Overall seroprevalence was 28.7% (95% CI: 26.9–30.5; Table 1), ranging between study centres from 16.7% (95% CI: 8.1–25.3) to 35.7% (95% CI: 25.5–46.0). Seroprevalence increased with age (Fig. 2).

Despite the associations detected and reported above, the random forest method showed that the potential risk factors considered in this study are not at all able to predict anti-HEV IgG ratio nor disease status (results are omitted). Indeed, the proportion of explained variance was practically zero in our prediction analysis, which in particular prevents us from ranking the risk factors.

The present population-based serological survey confirms the high HEV seroprevalence (29%) in the general population of the Netherlands, but provides no evidence for environmental transmission from pig farms to humans living in the vicinity of these farms.

Comparison of the seroprevalence found in this study with results from other European countries is difficult because of considerable differences in the studied populations, as well as the used assay for HEV antibody testing [54–56]. Recent studies using sensitive assays (like the Wantai assay we used) reported HEV seroprevalence varying between 16% and 52% [18, 57–61]. Our seroprevalence (29.7%) corresponds well with the 27% prevalence found in a large serological screening of Dutch blood donors [18]. Both studies used the Wantai assay to measure HEV specific IgG antibodies. In contrast, a population-based seroprevalence study in the general population in the Netherlands in 2006–2007, using a less sensitive assay for detection of prolonged IgG antibodies (MP Diagnostics, France), found a seroprevalence of only 3% [62]. Comparative studies suggest that the Wantai assay gives the most reliable estimate of the HEV seroprevalence, because it is more sensitive, it is positive in a higher proportion of proven infections and it remains positive for longer period post infection compared with the MP assay [56, 58, 59, 63].

In line with other serological studies, we found statistically significant higher seroprevalence in participants aged 50 years and older compared to 20–30 year olds [18, 44, 57, 62, 64]. This higher seroprevalence in older persons might be a result of age-dependent cumulative exposure, but can also be indicative of an age-cohort because of higher pressure of infection in the past. This age-cohort effect is also demonstrated in Denmark, the United Kingdom and the United States [60, 61, 65]. The HEV seroprevalence in blood donors as well as the high number of HEV RNA-positive donors found in serological screenings suggest an increasing HEV incidence in recent years in the Netherlands [18, 66, 67].

Proximity to pig farms and number of animals close to the home were used as proxy of potential environmental exposure to HEV. The lack of a positive association suggests that airborne spread of HEV is unlikely. Other environmental transmission routes cannot be excluded, for example through manure that may be transported to distant places or through surface water. In the Netherlands, HEV RNA has been detected in surface water samples [68]. Furthermore, a recent cross-sectional study suggests that consumption of tap water in France might be a risk factor for HEV infection [69]. Occupational exposure to pigs, which is a potential risk factor for seropositivity, was not investigated in our study since persons living or working on a farm were excluded [44, 47, 70]. Moreover, the present study was not designed to investigate the role of food-borne transmission. Presence of IgG antibodies is an indicator for HEV infection in the past and therefore suited to study exposure to HEV through the environment. Although we found no decreased risk for HEV seropositivity in participants who did not consume pork meat, this finding must be interpreted with caution because of the small numbers of participants on a diet without pork meat (n = 47, 2.0%). Moreover, some of these participants might have consumed pork meat in the past. A case-control study focussing mainly on food-borne transmission among patients with acute hepatitis E infection in the Netherlands is ongoing (study period: 2015–2017).

In conclusion, our study showed that the HEV seroprevalence in the general population in a livestock-dense area in the south of the Netherlands was not associated with living in proximity to pig farms. The higher seroprevalence among older participants supports the assumption of cumulative lifetime exposure to hepatitis E virus in the Netherlands, as well as an age-cohort because of higher pressure of infection in the past. Although it seems plausible that transmission is primarily food-borne, further research is needed to elucidate the exact sources and routes of infection.